Benchmarking machine learning models in lesion-symptom mapping for predicting language outcomes in stroke survivors

Several decades of research have investigated the neural connections between stroke-induced brain damage and language difficulties. Typically, lesion-symptom mapping (LSM) studies that address this connection have relied on mass univariate statistics, which do not account for multidimensional relati...

Full description

Saved in:
Bibliographic Details
Main Authors: Deepa Tilwani, Christian O'Reilly, Nicholas Riccardi, Valerie L. Shalin, Dirk-Bart den Ouden, Julius Fridriksson, Svetlana V. Shinkareva, Amit P. Sheth, Rutvik H. Desai
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-05-01
Series:Frontiers in Neuroimaging
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fnimg.2025.1573816/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849717450358128640
author Deepa Tilwani
Deepa Tilwani
Deepa Tilwani
Deepa Tilwani
Christian O'Reilly
Christian O'Reilly
Christian O'Reilly
Christian O'Reilly
Nicholas Riccardi
Valerie L. Shalin
Valerie L. Shalin
Dirk-Bart den Ouden
Julius Fridriksson
Svetlana V. Shinkareva
Svetlana V. Shinkareva
Amit P. Sheth
Amit P. Sheth
Amit P. Sheth
Rutvik H. Desai
Rutvik H. Desai
author_facet Deepa Tilwani
Deepa Tilwani
Deepa Tilwani
Deepa Tilwani
Christian O'Reilly
Christian O'Reilly
Christian O'Reilly
Christian O'Reilly
Nicholas Riccardi
Valerie L. Shalin
Valerie L. Shalin
Dirk-Bart den Ouden
Julius Fridriksson
Svetlana V. Shinkareva
Svetlana V. Shinkareva
Amit P. Sheth
Amit P. Sheth
Amit P. Sheth
Rutvik H. Desai
Rutvik H. Desai
author_sort Deepa Tilwani
collection DOAJ
description Several decades of research have investigated the neural connections between stroke-induced brain damage and language difficulties. Typically, lesion-symptom mapping (LSM) studies that address this connection have relied on mass univariate statistics, which do not account for multidimensional relationships between variables. Machine learning (ML) techniques, which can capture these intricate connections, offer a promising complement to LSM methods. To test this promise, we benchmarked ML models on structural and functional MRI to predict aphasia severity (N = 238) and naming impairment (N = 191) for a cohort of chronic-stage stroke survivors. We used nested cross-validation to examine performance along three dimensions: (1) parcellation schemes (JHU, AAL, BRO, and AICHA atlases), (2) neuroimaging modalities (resting-state functional connectivity, structural connectivity, mean diffusivity, fractional anisotropy, and lesion location) and (3) ML methods (Random Forest, Support Vector Regression, Decision Tree, K Nearest Neighbors, and Gradient Boosting). The best results were obtained by combining the JHU atlas, lesion location, and the Random Forest model. This combination yielded moderate to high correlations with the two different behavioral scores. Key regions identified included several perisylvian areas and pathways within the language network. This work complements existing LSM methods with new tools for improving the prediction of language outcomes in stroke survivors.
format Article
id doaj-art-d260ca7340c14979b081cf7be59f13ea
institution DOAJ
issn 2813-1193
language English
publishDate 2025-05-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Neuroimaging
spelling doaj-art-d260ca7340c14979b081cf7be59f13ea2025-08-20T03:12:39ZengFrontiers Media S.A.Frontiers in Neuroimaging2813-11932025-05-01410.3389/fnimg.2025.15738161573816Benchmarking machine learning models in lesion-symptom mapping for predicting language outcomes in stroke survivorsDeepa Tilwani0Deepa Tilwani1Deepa Tilwani2Deepa Tilwani3Christian O'Reilly4Christian O'Reilly5Christian O'Reilly6Christian O'Reilly7Nicholas Riccardi8Valerie L. Shalin9Valerie L. Shalin10Dirk-Bart den Ouden11Julius Fridriksson12Svetlana V. Shinkareva13Svetlana V. Shinkareva14Amit P. Sheth15Amit P. Sheth16Amit P. Sheth17Rutvik H. Desai18Rutvik H. Desai19Artificial Intelligence Institute, University of South Carolina, Columbia, SC, United StatesDepartment of Computer Science and Engineering, University of South Carolina, Columbia, SC, United StatesCarolina Autism and Neurodevelopment Research Center, University of South Carolina, Columbia, SC, United StatesInstitute for Mind and Brain, University of South Carolina, Columbia, SC, United StatesArtificial Intelligence Institute, University of South Carolina, Columbia, SC, United StatesDepartment of Computer Science and Engineering, University of South Carolina, Columbia, SC, United StatesCarolina Autism and Neurodevelopment Research Center, University of South Carolina, Columbia, SC, United StatesInstitute for Mind and Brain, University of South Carolina, Columbia, SC, United StatesDepartment of Communication Sciences and Disorders, University of South Carolina, Columbia, SC, United StatesArtificial Intelligence Institute, University of South Carolina, Columbia, SC, United StatesDepartment of Psychology, Wright State University, Dayton, OH, United StatesDepartment of Communication Sciences and Disorders, University of South Carolina, Columbia, SC, United StatesDepartment of Communication Sciences and Disorders, University of South Carolina, Columbia, SC, United StatesInstitute for Mind and Brain, University of South Carolina, Columbia, SC, United StatesDepartment of Psychology, University of South Carolina, Columbia, SC, United StatesArtificial Intelligence Institute, University of South Carolina, Columbia, SC, United StatesDepartment of Computer Science and Engineering, University of South Carolina, Columbia, SC, United StatesCarolina Autism and Neurodevelopment Research Center, University of South Carolina, Columbia, SC, United StatesInstitute for Mind and Brain, University of South Carolina, Columbia, SC, United StatesDepartment of Psychology, University of South Carolina, Columbia, SC, United StatesSeveral decades of research have investigated the neural connections between stroke-induced brain damage and language difficulties. Typically, lesion-symptom mapping (LSM) studies that address this connection have relied on mass univariate statistics, which do not account for multidimensional relationships between variables. Machine learning (ML) techniques, which can capture these intricate connections, offer a promising complement to LSM methods. To test this promise, we benchmarked ML models on structural and functional MRI to predict aphasia severity (N = 238) and naming impairment (N = 191) for a cohort of chronic-stage stroke survivors. We used nested cross-validation to examine performance along three dimensions: (1) parcellation schemes (JHU, AAL, BRO, and AICHA atlases), (2) neuroimaging modalities (resting-state functional connectivity, structural connectivity, mean diffusivity, fractional anisotropy, and lesion location) and (3) ML methods (Random Forest, Support Vector Regression, Decision Tree, K Nearest Neighbors, and Gradient Boosting). The best results were obtained by combining the JHU atlas, lesion location, and the Random Forest model. This combination yielded moderate to high correlations with the two different behavioral scores. Key regions identified included several perisylvian areas and pathways within the language network. This work complements existing LSM methods with new tools for improving the prediction of language outcomes in stroke survivors.https://www.frontiersin.org/articles/10.3389/fnimg.2025.1573816/fullaphasialesion-symptom mappingneuroimagingmultivariate analysisstrokemachine learning
spellingShingle Deepa Tilwani
Deepa Tilwani
Deepa Tilwani
Deepa Tilwani
Christian O'Reilly
Christian O'Reilly
Christian O'Reilly
Christian O'Reilly
Nicholas Riccardi
Valerie L. Shalin
Valerie L. Shalin
Dirk-Bart den Ouden
Julius Fridriksson
Svetlana V. Shinkareva
Svetlana V. Shinkareva
Amit P. Sheth
Amit P. Sheth
Amit P. Sheth
Rutvik H. Desai
Rutvik H. Desai
Benchmarking machine learning models in lesion-symptom mapping for predicting language outcomes in stroke survivors
Frontiers in Neuroimaging
aphasia
lesion-symptom mapping
neuroimaging
multivariate analysis
stroke
machine learning
title Benchmarking machine learning models in lesion-symptom mapping for predicting language outcomes in stroke survivors
title_full Benchmarking machine learning models in lesion-symptom mapping for predicting language outcomes in stroke survivors
title_fullStr Benchmarking machine learning models in lesion-symptom mapping for predicting language outcomes in stroke survivors
title_full_unstemmed Benchmarking machine learning models in lesion-symptom mapping for predicting language outcomes in stroke survivors
title_short Benchmarking machine learning models in lesion-symptom mapping for predicting language outcomes in stroke survivors
title_sort benchmarking machine learning models in lesion symptom mapping for predicting language outcomes in stroke survivors
topic aphasia
lesion-symptom mapping
neuroimaging
multivariate analysis
stroke
machine learning
url https://www.frontiersin.org/articles/10.3389/fnimg.2025.1573816/full
work_keys_str_mv AT deepatilwani benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT deepatilwani benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT deepatilwani benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT deepatilwani benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT christianoreilly benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT christianoreilly benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT christianoreilly benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT christianoreilly benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT nicholasriccardi benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT valerielshalin benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT valerielshalin benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT dirkbartdenouden benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT juliusfridriksson benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT svetlanavshinkareva benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT svetlanavshinkareva benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT amitpsheth benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT amitpsheth benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT amitpsheth benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT rutvikhdesai benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors
AT rutvikhdesai benchmarkingmachinelearningmodelsinlesionsymptommappingforpredictinglanguageoutcomesinstrokesurvivors